基于组合优化理论的红外图像边缘检测

Infrared image edge detection based on combinatorial optimization theory

边缘是红外图像的重要特征,检测好坏直接影响后续红外图像理解和识别,为了克服小波变换和蚁群优化算法各自存在的局限性,提出一种基于组合优化理论的红外图像边缘检测算法(WA-ACO)。该算法首先采用小波变换对红外图像进行处理,并根据高频子带相关性对边缘信息进行分类,然后采用蚁群优化算法从边缘信息中提取有用边缘,在消除噪声同时有效保护红外图像的弱边缘。实验结果表明,WA-ACO能够成功的提取红外图像边缘信息,抑制了噪声对边缘检测干扰,可以应用于红外图像处理和分析。

Edge is an important feature of infrared image and detection quality directly affect the subsequent infrared image understanding and recognition,in order to overcome the limitation of wavelet transform and ant colony optimization algorithm,a new algorithm is proposed infrared image edge detection algorithm WA-ACO based on combination optimization theory. The algorithm firstly use the wavelet transform for infrared image processing,and according to the high frequency sub-band correlation to classify the edge signal,and then use the ant colony optimization algorithm to extract the useful information edges from the edge,The experimental results show that WA-ACO can successfully extract edge information of infrared image,which can effectively suppress the interference of noise on edge detection,and can be applied to infrared image processing and analysis.